CA1287578C - Anti-human ovarian cancer immunotoxins and methods of use thereof - Google Patents

Abstract

Abstract of the Disclosure Immunotoxins comprising a cytotoxic moiety and an antigen binding portion selected from the group consisting of Fab, Fab' and F(ab')2 fragments of a monoclonal antibody, which binds to human ovarian cancer tissue, having one of the following capabilities are claimed: cytotoxic ID50 of about 10 nM or less against human ovarian cancer cells, retardation of human ovarian cancer tumor growth in mammals, or extension of survival of a mammal carrying a human ovarian cancer tumor. Antigens or epitopes to which the monoclonal antibodies bind are identified and characterize the immunotoxins. In a preferred embodiment an immunotoxin comprising at least an antigen binding portion of a monoclonal antibody, which binds to human transferrin receptor, but does not block binding of transferrin to the receptor, is described and claimed. Immunotoxin comprising the (F(ab')2 region of the antitransferrin monoclonal antibody are also claimed.
Methods of killing human ovarian cancer cells, retarding the growth of human ovarian cancer tumors in mammals and extending the survival of mammals carrying human ovarian cancer tumors are claimed.

Description

$/~

ANTI-HUMAN OVARIAII CANCER IMhlUNOTOXINS
AND l~lETHODS OF USE THEREOF

This invention is in the fields of immunology and cancer diagnosis and therapy. ~ore particularly it concerns murine monoclonal antibodies active against human ovarian cancer, hyhridomas that produce those antibodies, immunochemicals made from those antibodies, and diagnostic and therapeutic methods that use those immunochemicals.
A~ong gynecological malignancies occurring in American women, ovarian cancer most frequently causes death. The malignancy remains confined to the peritoneal cavity during practically its entjr~ clinical course. Characteristically the tumor disseminates throughout the peritoneal cavity producing acites and tumor foci on multiple peritoneal surfaces. The disease cannot be effectively cured lS surgically and chemotherapy is increasingly the primary treatment.
~ecause ovarian t~mors generally remain in the peritoneal cavity, c'nemotherapeutic agents may be administered system;cally by intravenous injection or by direct infusion into the peritoneal cavity thus by-passing the circulatory system as the route for initially exposing the tumor to the chemotherapeutic agent.
The use of monoclonal antibodies against antigens associated with cancerous ovarian tissues has been reported to only a limited extentc An antibody to human transferrin receptor linked to Pseudomonas exotoxin has been reported to have cytotoxic activity in ` 25 certain human ovarian cell lines. Pirker et al., "Anti-transferrin receptor antibody linked to Pseudomonas exotoxin; a model immunotoxin in human ovarian carcinoma cell lines", Cancer Res. 45:751-757 (19~5). Anti-transferrin monoclonal antibodies that inhibit the binding of transferrin to the transferrin receptor are the subject of U.S. Patent 4,434,156. The anti-transferrin monoclonal antibodies of the present invention are different from those disclosed in U.S.
Patent 4,434,156. Although the anti-transferrin antibody of the present invention binds the transferrin receptor, it does not inhibit the binding of transferrin to the transferrin receptor. Schlom et .~ '- ~

: , '. '~' , ` . ' ' , ." ' " .: ' ~l~87~

al., U.S. Patent 4,522,918 discloses a method of producing monoclonal antibodies against certain human breast cancer tumors using soluble extracts of human breast cancer.
A principal aspect of the invention concerns murine monoclonal antibodies that:
(a) bind human ovarian cancer tissue frozen sections;
(b) are IgGs or Ig~s;
(c) when bound to a cytotoxic moiety, have an ID50 of 10 n~
or less against at least one ovarian cancer cell line selected from the group consisting of OVCM-2, WCAR-3, OV~ M-4, WCAR-5 or A1847;
or when bound to a cytotoxic moiety extend the survival time of mammals carrying human ovarian tumorsi or when bound to a cytotoxic 15 protein retard the rate of growth of human ovarian tumors carried by such mammals.
Preferred embodiments of these antibodies are those designated 2G3, 9C6, 33F8, 44B2, 44F4, 120~7, 200F9, 204F4, 219F3, 245E7, 260F9, 266B2, 280Dll, 317G5, 369F10, 388D4, 421E8, 451C3, 20 454A12, 454Cll, 650E2, 788G6, 871E3, and functional equivalents thereof.
T~e murine x murine hybridomas that produce the above described antibodies and progeny of those hy~ridomas are other aspects of the invention.
Another aspect of the invention relates to immunotoxins that are conjugates of (a) the above described monoclonal antibodies, and ~b) a cytotoxic moiety.
Another aspect of the invention concerns methods of extending the survival time of mammals bearing human ovarian tumor cells bX administering to such mammal an amount of an immunotoxin described above effective to extend the life of such mammal.

~` :
.

, ~ . . ~. . - ; . .
. - : .
~ .. - , .

: . . . .. ~ .. . . .
:~,,; . . . . .
: ,. ...
- .: .
.~. . ~ .
.. . : '-: ` -; ~, '. :. ,, :

~37 Yet another aspect of the invention concerns rnethods of killing hurnan ovarian tumor cells by contacting such cells with a cytocidally effective amount of the imrrunotoxin descr;bed above.
A further aspect of the invention concerns methods of 5 retarding the rate of growth of human ovarian tumor cells carried by a mammal by administering to such mammal a tumor cell growth-retarding amount of the immunotoxin described above.
As used herein, the term "monoclonal antibody" means an antibody composition having a homogeneous antibody population. It is .10 not intended to be limited as regards the source of the antibody or the manner in which it is made.
As used herein the term "antigen binding portion of a monoclonal antibody" means the portion of the monoclonal antibody that binds an antigen to which t'ne monoclonal antibody is specific. In 15 general, such antibody binding portions of the monoclonal encompass the Fab, Fab' and F(ab')2 regions or fragments of the immunoglobin ~;molecule. Fab, Fab' and F(ab')2 regions of an im~unoglobin may be generated by enzymatic digestion of the monoclonal antibodies using techniques well known to those skilled in the art. Fab fragments may 20 be generated by digesting the monoclonal antibody with papain and contacting the digest with a reducing agent to reductively cleave disulfide bonds. Fab' fragrnents may be obtained by digesting the antibody with pepsin and reductive cleavage of the fragment so produced with a reducing agent. In the absence of reductive cleavage, 25 enzymatic digestion of the monoclonal with pepsin produces F(ab')2 fragments.
As used herein with regard to the monoclonal antibody-producing hybridomas of the invention the term "progeny" is intended to include all derivatives, issue, and offspring of the parent hybridorna that produce the monoclonal anti-human ovarian cancer antibody produced by the parent, regardless of generation or karyotypic identity.
As used herein with respect to the exemplified r~rine monoclonal antibodies against human ovarian cancer, the term .

.. , .. ~ .. ~
. . ,: . , ~.-.: , . . ., , ' , ' . .... . . . .

. .
- ., ... : . .

~L~8~571~3 "functional equivalent" rneans a monoclonal antibody that: (a) binds to the same antigen or epitope as an exemplified monoclonal antibody as determined by immunoprecipitation or sandv~ich assa~; (b) binds human ovarian cancer tissue frozen sect~ons; (c) has a selectivity 5 equal to or less than 0.11; (d) has a G or M isotype, and (e) when conjugated to a cytotoxic moiety forms an immunotoxin which (i) extends the survival of a mammal bearing human ovarian cancer cells when administered to such mammal or (ii) retards the growth of human ovarian cells in a mammal bearing such cells when administered to such a mammal or ~iii) is cytotoxic to human ovarian cancer cells when such cells are contacted with the immunotoxin.
As described above, the term "functional equivalent" as used herein inclu~es five criteria. The first of these criteria9 binding to the same antigen or epitope as an exemplified monoclonal antibody 15 may be demonstrated by experiments which show crossblocking of an exemplified monoclonal antihody by the functionally equivalent monoclonal antibody. Crossblocking occurs as a result of an antibody binding to the same epitope on an antigen as that bound by one of the exemplified antibodies, or as a result of an antibody hinding to a 20 different epitope which is so closely situated on the same antigen that binding of an antibody to one epitope blocks the binding of an antibody to the second epitope. Crossblocking thus is one of the criteria by which one can determine that a functionally equivalent monoclonal antibody binds to the same antigen or epitope as an 25 exemplified monoclonal antibody.
So-called "sandwich" assays are another method for determining whether an antibody binds the same antigen or epitope. In these assays, a first monoclonal antibody is bound to a support, for example, the surface of a titre plate well. After treatment to 30 prevent nonspecif;c binding, a highly solubilized antigen preparation is added to the bound antibody. Subsequently, a second antibody, having a detectable label, for example, a fluorescent dye, is added.
If the second antibody binds to the antigen, a different epitope specificity or multiple copies of the same epitope on the same antigen 35 is indicated. If the second antibody fails to bind, either the same .
,. :, .

. . :

2~37~

epitope specificity or different antiyen specificity is indicated~
The results of both the crossblocking and sandwich assay are further defined by a second series of test~ such as immune precipitation or Western blotting to show that the antigen bound by both antibodies has the same molecular WPi ght.
The immunotoxins according to the invention are conjugates of the monoclonal antibody and a cytotoxic moiety. The cytotoxic moiety of the immunotoxin may be a cytotoxic drug or an enz~matically active toxin of bacterial, fungal or plant origin, or an enzymatically active polypeptide chain or fragment ("A chain") of such a toxin.
~` Enzymatically active toxins and fragments thereof are preferred and are exemplified by diphtheria toxin A fragment, nonbinding active fragments of diphtheria toxin, exotoxin A (from Pseudomonas aeruginosa), ricin A chain, abrin A chain~ modeccin A chain, alpha-sarcin, certain Aleurites fordii proteins, certain Dianthin proteins, Phytolacca americana proteins (PAP, PAPII and PAP-S), ~omordica charantia inhibitor, curcin, crotin, Saponaria officinalis inhibitor, gelonin, mitogellin, restrictocin, phenomycin, and enomycin. Ricin A, chain, Pseudomonas aeruginosa exotoxin A and PAP are preferred.
Conjugates of the monoclonal antibody and such cytotoxic moieties may be made using a variety of bifunctional protein coupling agents. Examples of such reagents are N-succinimidyl-3-(2-pyridyldithio) propionate (SPDP)a iminothiolane (IT), bifunctional derivatives of imidoesters such as dimethyl adipimidate HCl, active esters such as disuccinimidyl suberate, aldehydes such as glutaraldehyde, bis-azido compounds such as bis(p-diazoniumbenzoyl)-ethylenediamine, diisocyanates such as tolylene 2,6-diisocyante, and bis-active fluorine compounds such as 1,5-difluoro-2,4-dinitrobenzene.
The enzymatically active polypeptide of the immunotoxins according to the invention may be recombinantly produced.
Recombinantly produced ricin toxin A chain (rRTA) may be produced in accordance with the methods disclosed in PCT W085/03508 published August 15, 1985. Recombinantly produced diphtheria toxin A chain and non-binding active fragments thereof are also described in PCT
~085/03508 published August 15, 1985.

.
. .

~- ~- - . .. ....

. ;: ~, .. : : , ': .~
, . .

~ .
~2l3~5~
~ 6 _ When used to kill human ovarian cancer cells in vitro for diagnostic purposes, the conjugates will typically be added to the cell culture medium at a concentration of at least about I0 n~. The formulation and mode of administration for in vitro use are not critical. Aqueous formulations that are compatible with the culture or perfusion medium will normally be used. Cytotox;ci~y may be read by conventiona1 techniques suc~ as dye exclusion or inhibition of colony formation in a clonogenic assay to determine the presence of an ovarian cancer tumor that is susceptible to treatment with the immunotoxin of interest.
When used _ vivo for therapy; the imnunotoxins are admin;stered to the patient in therapeutically effective amounts li.e., amounts that eliminate or reduce or retard the increase of the patientls tumor burden). They will normally be administered parenterally, preferably intraperitoneally (IP). The dose and dosage regimen will depend upon the nature of the cancer (primary or metastatic) and its population, t~e characteristics of the particular ~` immunotoxin, e.g., its therapeutic index, the patient, and the, patient's history. The amount of immunotoxin administered (IP) will typically be in the range of about O.OI to ahout I00 mg/kg and preferably between O.OI mg/kg and 10 mg/kg of patient weight.
For parenteral administration the immunotoxins will be forl~lated in a unit dosage injectable form (solution, suspension, emulsion) in association with a pharmaceutically acceptable parenteral vehicle. Such vehicles are inherently nontoxic and nontherapeutic.
Examples of such vehicles are water, saline, Ringer's solution, dextrose solution, and 5% human serum albumin. ~onaqueous vehicles such as fixed oils and ethyl oleate may also be used. Liposolnes may be used as carriers. The vehicle may contain minor amounts of additives such as substances that enhance isotonicity and chemical stability, e.g., huffers and preservatives. The immunotoxin will typically be formulated in such vehicles at concentrations of about O.OI mg/ml to I00 mg/ml.

.
. .
. .. . . .

~a28'-~S7~

Cytotoxic radiopharmaceuticals for treating ovarian cancer may be made by conjugating high linear energy transfer (LET~ emitt;ng isotopes (e,g., Y, Pr) to the antibodies. The term "cytotoxic mo;ety"
as used herein is intended to include such isotopes.
The antibody producing fusion partners that are used to make the hybridomas of this invention are generated by immunizing mice with live human breast cancer cells or membrane extracts made therefrom.
The mice are inoculated intraperitoneally with an immunogenic amount of ~he cells or extract and then boosted with similar amounts of the immunogen. Spleens are collected from the immunized mice a few days ` after the final boost and a cell suspension is prepared therefrom for use in the fusion.
~ybridomas are prepared from the splenocytes and a murine tumor partner using the general somatic cell hybridization technique of Kohler, ~. and ~ilstein, C., Nature (1975) 256:495-497 as modified - by Buck, D. W., et al, In Vitro (1982) 18:377-381. Available murine myeloma lines, such as those from the Salk Institute, Cell Distribution Center, San Diego, California, USA, may be used in the' hybridization. Basically, the technique involves fusing the tumor 20 cells and splenocytes using a fusogen such as polyethylene glycol.
After the fusion, the cells are separated from the fusion medium and grown in a selective growth medium, such as HAT medium, to eliminate unhybridized parent cells. The hybridomas are expanded, if desired, and supernatants are assayed for anti-human breast cancer activity by 25 conventional immunoassay procedures (e.g., radioimmunoassay, enzyme immunoassay, or fluorescence immunoassay) using the immunizing agent (breast cancer cells or membrane extract~ as antigen. Positive clones are characterized further to determine whether they meet the criteria of the antibodies according to the invention.
Hybridomas that produce such antibodies may be grown ln vitro or in vivo using known procedures. Preferably the hyhridomas are maintained as ascites in mice. The monoclonal antibodies may be isolated from the culture media or body fluids, as the case may be, by conventional immunoglobulin purification procedures such as ammonium .

.:.
.. . .
:' . ,, - . ~:

... .- ~ -. . . , . - . , .
.: . , ~ . , .

7~7~3 sulfate precipitation, gel electrophoresis, dialysis7 chromatograpny, and ultrafiltration, if desired.
` The important characteristics of the monoclonal antibodies are (I) their immunoglobulin class, (2) thelr ability to bind hu~an ovarian cancer tissue, (3) their selectivity as defined further hereinbelow (4) their usefulness in making effective anti-hunan ovarian cancer immunotoxins which are either cytotoxic to human ovarian cancer cells, or extend the survival of mammals carrying human ovarian cancer cells, or retard the growth of human ovarian cancer cells in animals bearing such cells. T~e monoclonal antibodies :suitable as immunotoxins according to the invention were initially identified as monoclonal antibodies within a group of anti-breast cancer monoclonal antibodies.
In selecting the claimed antibodies, approximately 22,000 gro~ing hybridoma cultures were initially screened against the im~,unizing breast tumor membranes or cell line, a panel of seven normal tissue membranes, a fibroblast cell line and a breast tumor frozen section. Clones that reacted with the neoplastic materials, but not ~he normal materials, were identified in this initial screen 20 and chosen for isotyping and additional screening for selectivity and range. The additional screening involved: sixteen normal tissue sections, five normal blood cell types, eleven nonbreast neoplasm sections, twenty-one breast cancer sections and fourteen breast cancer cell lines. In the additional screening, a number of monoclonal 25 antibodies bound ovarian carcinoma tissue sections strongly but did not appear to bind to normal ovarian tissue sections.
For purposes of this patent application, specificity and selectivity are used interchangeably and are defined as the sum of the number of substructures stained in sixteen normal tissue frozen sections and the number of blood cell types bound, divided by the sum of the total number of substructures bound by any of the rnonoclonal antibodies in all the tissues on which the monoclonal antibodies were tested and five blood cell types tested. 123 Substructures and five blood cell types were counted in the tests. Antibodies were deemed to , ~ "

, . - : , .
- - . , ., ~ ., . : ..

:. ~
- - . . ,- ~. . : ., ,, , :
.

~2B7:~7~

be appropriate candidates for ovarian cancer ;mmunotoxin purposes if they have a selectivity equal to or less than 0.11 and bound to human ovarian cancer tissues.
Antibodies produced by one of the hybridomas were found to recognize a Z00 K dalton antigen. Antibod~es of two of the hybridomas bound to a 42 K dalton antigen. Four bound to one or more ~igh molecular weight mucins (HMW) and two bound to transferrin receptors in the form of a 95 K dalton antigen. Two bound to the same epitope of a 55 K dalton antigen. All antigen weights mentioned herein were ~lO determined by sodium dodecyl sulfate (SDS) polyacrylam1de gel ``electrophoresis under reducing conditions using procedures known in the art.
Further details of the characterization of these antibodies are provided in the examples below.
The immunochemical derivatives of this ;nvention that are of prime importance are conjugates of the monoclonal antibodies and a cytotoxic agent.
Fresh postsurgical human breast cancer tissue and a variety of normal tissues were used to prepare membrane extracts by homogenization and discontinuous sucrose gradient centrifugation.
Human breast cancer cell lines were obtained from the Breast Cancer Task Force, the American Type Culture Collection (ATCC), and from Dr.
Jorgen Fogh at ~emorial Sloan Kettering. The cells were maintained and passaged as recommended by the Breast Cancer Task Force, the ATCC
and Dr. Fogh. For immunizations, either membrane extract containing 100 ~g of protein (Lowry assay) or ten million live breast cancer cells were inoculated intra-peritoneally into five week old Balb/c mice. The mice were boosted identically twice at monthly intervals.
Three days after the 7ast boost, the spleens were removed for cell fusion.
Somatic cell hybrids were prepared by the method of Buck, ~.
W., et al, supra, using the murine myeloma line Sp-2/0/Agl4. All hybrodima cell lines were cloned by limiting dilution. Half of the fusions employed splenocytes from mice immunized with breast cancer .

~2~7~

membrane extracts and half used splenocytes from mice immunized with live breast cancer cell lines. Eighty-three thousand Four hundred twenty-four wells were genera~ed from those fusions~ of which 22,4~9 exhibited hybridoma growth.
Hybridoma supernatant was assayed for reactive antibody in either a solid phase enzyme-linked immunosorbent assay (ELISA) with the immunizing breast cancer membrane extract or an indirect immunofluorescence assay with the immunizing breast cancer cell line. For the solid phase membrane ELISA, 40 ~l of 0.1 mg/ml breast ~ 10 cancer membrane protein were placed in polyvinyl chloride (PVC) `` microtiter wells for 12 hours at 4C. The extract was aspirated and the wells washed with phosphate buffered saline (PBS) containing 1%
bovine serum albumin ~BSA). The wells were then incubated with 45 ~l of a 1:10 dilution of hybridoma supernatant. The diluent was media 15 with 25 mM of a buffer, 10% bovine serum, and 0.1% sodium azide.
After 30 minutes at room temperature, the wells were again washed and incubated 45 minutes at 37C with a 1:200 dilution of peroxidase i~ conjugated goat anti-mouse IgG. The diluent was PBS. The wells were, then washed with PBS and reacted with 200 ~l of 1,2-azino-di(3-ethylbenzthiazoline sulphonic acid) in 0.1 M sodium citrate buffer pH
4.2 for 30 minutes at room temperature. Optical density was measured at 405 nm on a MicroElisa Reader. For each experiment a positive control, anti-beta 2 microglobulin at 5 ~g/ml, was reacted with normal human kidney membrane. This gave an optical density of 1.0 i 0.1 (standard deviation~. The background was O + 0.1 optical density units (O.D.) using media without mouse monoclonal antibody. Wells that gave a reaction on the breast cancer membrane extract of greater than 0.7 O.D. were saved.
For the indirect immunofluorescence cell line assay 100,000 breast cancer cells of the immunizing cell line were placed overnight with appropriate media in each chamber of a set of eight chambered slides. Similarly, 100,000 fibroblast cells from cell line CC95 were incubated overnight in chambered slide wells. The cells were washed with PBS containing 1~ BSA. The wells, both breast cancer and fibroblast, were incubated for 30 minutes at 4C with 1:10 dilutions - : . ; ~ , , . . .:
, ' :. : ' ': '' ~ , :, ,.
. ,:

7~

of hybridoma supernatant. The cells ~ere aga;n washed and incubated 30 minutes at 4C with a 1:50 dilution of fluorescein isothiocyanate (FlTC)-conjugated goat F(ab')2 anti-mouse Ig. The cells were washed three times, fixed in 1.5% formaldehyde in PBS for five minutes, chambers removed and rinsed in PBS. The slides were then mounted in a composition containing polyvinyl alcohol, glycerol, buffers and a preservatjve and examined with a fluorescencr- microscope. Hybridoma wells showing strong fluorescent binding to the breast cancer cells but no fluorescent binding to fibroblasts were saved. Five thousand one hundred fifty-six hybridoma wells revealed breast cancer reactivity in the initial screen.
Supernatants from the 5156 positive wells were then tested in solid phas2 ELISA with seven normal tissue membrane extracts (liver, lung, Golon, stomach, kidney, tonsil, and spleen). Any supernatant giving an ELISA O.D. greater than 0.3 was discardet~. ~ e thousand one hundred one of the supernatants were found to be unreactive with the normal tissue extracts.
The 1101 hybridoma supernatants were tested on frozen sections of human breast carcinoma tissues. Six micron sections were ~ attached to slides, fixed 10 minutes in acetone at 4C, dried 10 minutes at room temperature, washed with PBS, blocked with horse serum and incubated 20 minutes at room te~perature with 100 ~l neat hybridoma supernatant. The slides were washed with PBS, and finally incubated 20 minutes at 37C with a 1:50 dilution of peroxidase conjugated rabbit anti-mouse Ig, washed again with PBS, and finally incubated 7.5 minutes at 37O with 0.5 mg/ml diaminobenzidine in 0.05 M Tris buffer pH 7.2 containing 0.01% hydrogen peroxide. The slides were stained with hematoxylin, dehydrated and mounted in a medium containing 35.9% methyl/n-butylmethacrylate copolymer, 7.1% butyl benzyl phthalate, and 0.3% 2,6-ditertbutyl-p-cresol. One hundred twenty-four wells yielded breast cancer selective binding and were cloned.
Immunoglobulin class and subclass of the monoclonal breast cancer selective antibodies were determined by an immunodot assay .
, . . :
- , . . , . :
-.
. . . .

essentially the same as that described in McDougal et al. J. Immunol.
Meth. 63:281-290 (1983). Antibodies were also internally labeled by growing 2-3 x 106 hybridoma cells for four hours in methionine-free medium containing 0.2 ~Ci 3~S methionine. 35S-labeled antibodies were 5 immunoprecipitated with fixed staphylococcus A cells, or with fixed staphylococcus A cells precoated with rabbit anti-mouse immunoglobulin, and the immunoprecipitates were analyzed by SDS-PAGE
to determine antibody light and heavy chain mobility, lack of extra chains, and the ability of each antibody to bind staphylococcal protein A.
The antibodies were expanded in vivo. Balb/c or F1 (C57~/6 x Balb/c) mice were primed with 0.5 ml pristane intraperitoneally (ip) and after 10-14 days inoculated with one million log phase hybridoma cells in PBS. Ascites fluid was stored at -70C and thawed and 15 filtered thrGugh a 0.8 micron filter unit before further purification.
Some IgG antibodies that bound staphylococcal protein A were purified by affinity chromatography on protein A-chromatographic resin containing either agarose, dextran and/or acrylamide ~ith pH step gradient elution. IgG antibodies that did not bind protein A were 20 precipitated by addition of ammonium sulfate to 40~ saturation at 0C.
or by binding to DEAE or Affigel~ (Biorad, Richmond, California).
Alternatively, IgG antibodies were purified by chromatography using a Sephacryl*S-200 column, followed by DEAE cellulose as described. The precipitates were redissolved in PBS, dialysed to 20 mM Tris p~ 7.2 25 and chromatographed on a 1.6 x 50 cm column of diethylaminoethyl cellulose (OEAE) eluting with a 105 liter 0-600 mM NaCl gradient at 4C at a flow rate of 1 ml/min. In each case, column fractions were monitored by SDS-PAGE and the purest antibody fractions were pooled, concentrated to 1-3 mg/ml, dialysed to PBS/0.02% NaN3~ and stored at IgM antibodies were purified by gel filtration material on a 2.6 x 40 cm column of Sephacryl S-300 or other gel filtration or resin containing agarose, dextran and/or acrylamide, eluting with PBS/0.01%
sodium azide at room temperature at a flow rate of 1 ml/min.

*Trade Mark - . ~ .
'' - ~ . , ' ' .

.
.

~2~3~S~

In order to evaluate their selectivity, the purified antibodies were tested by immunoperoxidase section staining on sections of sixteen normal tissues, and by immurlofluorescent cell sorting on five blood cell types. Im~unoperoxidase staining was performed as above except that known dilutions of purified anti~odies in PBS in the range of 1-40 ~g/ml were used instead of hybridoma supernatants. The pure antibodies were first titrated to find the minimal concentration giving strong immunoperoxidase staining on breast cancer sections and then used at the concentration for the normal tissue tests. No normal ovarian tissue showed detectable binding.
Peripheral blood cells (platelets, lymphocytes, red blood cells, granulocytes, and monocytes) were prepared by centrifugation using a medium which separates monocytes from polymorphonuclear leukocytes. The cells were reacted with antibody at the optimal concentration determined above for 30 minutes at 4C, washed, reacted with a 1:50 dilution of fluorescein isothiocyanate-conjugated goat anti-mouse Ig for 30 minutes at 4C, washed again and examined in a cell sorter. The wash buffer and diluents were PBS with 1% gelatin and 0.02% sodium azide. The cell sorter was equipped with a 76 micron nozzle and a one watt argon ion laser at 483 nm. An 80 mm confocal lens was used on the optical rail assembly for focusing. Okher filters used were a 515 nm interference filter and a 515 nm absorbance filter (for scattered laser liyht) and a neutral density 1.5 filter for forward angle light scatter. Contour plots of log fluorescein fluorescence versus forward angle light scatter were used for sample analysis.
The binding behaviors on normal tissue sections of the antibodies useful in the immunotoxins according to the invention are reported in Table 1 below. The following abbreviations are used to denote structures bound by the antibodies: Ac, acini; G, glands; T, tubules; ~, ducts; L, lumen; W, sweat glands; E, epithelium; S, sebaceous glands; Gr~ granulocytes; Mk, megakaryocytes; M, macrophage;
Ly, lymphocytes; Bl, Basal layer; Fe, focal epithelium; A, aveolar lining cells; B, Bowmanls capsule; Mu, nuscle; I, islets; X, '. : . ~ '' ' . : ' .-: ', ' :
. ' .

~2~7S78 ganglia/nerve; V, blood vessel; and H~ hair follicle. Selectivity was quantified as described hereinabove. The binding behavior of the antibodies on peripheral hlood cells ~s reported in Table 20 The selectivity of the monoclonal antibodies is set out in Table 3.

; ., : ~ :
::

: ::
:

', ~ , ' - , ~ . .

`.,: : - .

. , ,' , ' . ' , : : '. ' . ' ~7~7~

o i g ~ ~ ~W~ ~ ~
~D ~ O _ O O O O O O _ ~ _ O O O N O o O _ O O O
!~ ' O~ ~
O ~ ~ O _ O ~ ~ O N _ ~--O N -- O ~ N O O
l.-J i ~i O
~: SF j o~_o~ooo--ooo~ooOo~--oOOO
S ~ , S
VI C O -- O --~ N IC ~ ~C N ~`J _ O _ N O O O _ O O _ C OooOOOoOCOOOO~OOOOOOOO~

~_ OC:OO~OOOOOOOOOOOOOOOOOC
C ~ C 0 0 J 000--000000~0~00_00--~00 O J
O ~ - . J
L~- O -- j -- -- -- O -- C O ~ N e~ O O O-- O N 0 -- O -- --.¢ , Z
iOOCOOOO~OOOOO~OOO--OOOOO
S
~_ O ~ ~ _ ~ ,~ ~ J ~ ~ ~ _ t~ J ~ ~
~ ~ o ~ c N O ~-- O ~ O -- _ _ O O O O O --,coco~oo~oooo~~o--oo~o~o~

Y j ~J O -- O _ _ 1~. ~ _ O _ _ N ~ O _ _ O O _ ~ N O
lz~
~Soooooc,oo5ooc,ooooi~iro ' oo 1.~ ~ ~ N ~ _ ~ _ O N N O ~ ~-- O O N _ O O _ O O --O

~ S ~ U U U U ~ U ~ ~ U ' ~ S i ~ 5 --o _ o _ _ _ 5 o ~S o ~5 ~5 o o _ :S o N ~ ~ ~ j ~ ~ O O N O O _ ~ ~ N ~ (~ O C -- C O C ~ -- O C
j ~ ~J N ~r o o ~ ~ ~,~ O .0 8 '' o ~ Y o ~ _ ' ~ ~ ~ ~ ~ o ~ ~ Q ~ ~ S ~ ~

,,,~t ~ . .
~Z~7~7~

BLOOD CELL BINDINGS OF OVARIAN MABS
MABRBC PLATELET LYMPHOCYTE GRANULOCYTE MONOCYTE
_________ _________________________________________________ 20 454Cll O O O O O

`.. ,.: :

`: - `: - : ' ~: . .. . .

- ~: . . , :
~ ' , ~ - ' ' .
... . . . . .

`; ^ ` ` ~`:
31Z~757~
f TISSUE SELECTIVITY OF OVARIAN MA~S
NORMAL TISSUE
SUBSTRUCTURES
BOUND/NOR~AL
TISSUE SUB-MAB BLOOD CELLS STRUCTURES AND SELECTIVITY
BOUND BLOOD CELLS
__________ ________________________________ _ _______ 3 33F8 0 / 57 / 128 0.055 '` 5 44F4 1 / 512 / 128 0 094 7 200F9~ 0 / 53 / 128 0 023 12 266B2 0 / 59 / 128 0.070 13 2~0D11 1 / 512 / 12~ 0.094 14 317G5 0 / 56 / 128 0.047 19 454Al2 0 / 54 / 128 0 031 20 454C11 0 / 510 / 128 0 07~
21 650E2 0 / 58 / 128 0.063 22 788G6 0 / 52 / 128 0.016 23 871E3 0 / 511 / 128 0.086 The~antlbodies were tested by immunoperoxidase staining on eleven non-breast ma1ignancies. The results for the antibodies are reported in Table 4 below.

:: :
: ~ :

, :
. : ~ ,:.

, : ' . . . ; , . . .. : - , . , .' : , ' :
, . . . . . . . . . . .

7S~

,~z .
o ~ I N ~ --I ~ N ~ ~ N ~ 1 N ~ ~t N N ~

~ ~ I O O O O O O O O O O O O O O O ~ 1 ~4 N O O O C
:E Z
V~ I
'` o~IO~OOOOOOOO~--'000~ JOOOO

Y
V~ ~ I o O o ~ o O o o o o o o o o o o o _ ~ o o o o Z
C~ I
l_ I T
Cl~ j~OOOO~OOOOO~OOOOOC~ O~O
LL. ~ I
O I S I
--~ --~ ~ ) C~J --~ ) ~0 j " , _ Y I ~J ~ ~ O _I ~ O ~ _ ~ ~ _ _ _ O N ~ J O O

I_ Y~
LLJ ` I
I
I ~ C~ ~ ~ ~ ~ ~--~ ~ N O
V~ I
~: I
c~ O ~ i ~ O O O _~ N O C~l _I N ~ O ~ O ~ J O O
O C~
Z
C~ j I O O _ O O O ~ --O C~J O . ~ O ~ ~ ~ O O C~J N ~1 I

Z
Jo I ~--~ O O O O O O O O O O _ O ~ ~ _ O O C~J O O

I ~ ~ LL a~ Ll_ ~ O ~ m o ~ o 1~ r O 00 ~

.

, .~,.

: -, ' ' . :
' ~ ' ',`' ' .' , '' ' . ' , - ', ',' , . : ' : ' -` , 37~7i~

Several of the antibodies were iodinated and tested for binding to MCF-7, CAMAl, SKBR3 or ZR7530 cells. The antibodies were labeled wit~ 125I using chloramine T or Iodogen~ to a specific activity of approximately 5-10 ~Ci/~g. To determine 5 immunoradiochemical purity, 100,000 cpm of two of the labeled antibodies in 0.5 ml fetal calf serum was serially absorbed with five aliquots of target cells for 15 minutes at 0C (generally 4,000,000 cells per aliquot), and the remaining radioactivity in the supernatant after each absorption was determined.
For measurements of association constants, known concentrations of labeled and unlabeled monoclonal antibodies were incubated with target cells in fetal calf serum for 15 minutes on ice. Aliquots of the cell/antibody mix were then counted in a gamma counter or filtered through ~icrofold filter plates (V & P Scientific) 15 and the filters counted. To account for unbound antibody retained in liquid on the filters, controls containing the same concentrations of antibody but no cells were done in pafallel. Association constants and antigen copy number per target are calculated from the affinity test results and are reported in Table 5 below.

. - , . . . .

- ~ Z&1~7~
-2~-AFFINITY AND ANTIGEN COPY ~U~BEK OF OVARIAN MABS
MA8 n Ka CELL
LlNE
________ _ ____ ________ __ ______ _____ ___ 1 2G3 3700000 9.1x106 t~CF7 5 44F4 2100000 5.3x~o6 ~CF7 6 120H7 210000 2x10 MCF7 8 204F4 3200000 8.0x106 ~CF7 11 260F9 310000 5.6x107 MCF7 12 266B2 80000 2.7x108 MCF7 13 280D11 390000 8.8x106 ~CF7 14 317G5 3200000 1.6x106 CAMA1 1~ 451C3 400000 4X108 ~CF7 19 454A12 470000 1.2x108 MCF7 20 454C11 390000 4.8x107 ZR7530 22 788~6 .
"In order to idertify the antigens recognized by the monoclonal antlbodies, immunoprecipitation of the antigens was carried out according to the following method. Eight mm diameter polystyrene balls (Precision Plas~ic Ball Co.) were covered with 10% fuming nitric 5 acid in glacial acetic acid and were incubated for three hours in a 50C water bath. Following the acid treatment, the balls were rinsed three times with distilled water, covered with 1% sodium dithionite in 0.1 M NaOH and incubated three hours in a 50C water bath. The balls were again rinsed three times with distilled water, covered with 0.1~ -10 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide (EUAC), 0.2% suberic acid (suberic acid dissolved in dimethylformamide) and incubated overnight at room temperature. The balls were rinsed three times with distilled water, and marked for identification.

.~
.
.

.
. . .
. . ~: : : . . . : - .
. . .. .

Purified monoc'lonal antibodies were diluted 0.2 mg/rnl in 2-(N-morpholino)ethane sulfonic acid buffer, and the previously treated and marked polystyrene balls were placed in individual tubes and covered with 450 microliters diluted antibody and 50 microliters of fresh 1~ EDAC. Tubes were capped and incuhated at 25C for 24 hours. Following this incubation, the balls were rinsed t~ice with PBS and were either used fresh or were stored for several days a~ 4C
before use.
Freshly labeled target cell extracts were prepared from human breast cancer cell lines labeled with 125-I by the lactoperoxidase method of ~archalonis, J., "An Enzymic Method for the Trace Iodination of Immunoglobulins and other Proteins", Biochem. J.
113:299-305 (1969), or with 35-S by growth in 35-S methionine. The labeled cells were dissolved in solubilization buffer (1% (v/v) Triton X-100,* 150 mM NaCl, 5 mM EDTA, 25 mM Tris-HCl, pH 7.5). Four parts of labeled extract were mixed in a vessel with one part solubilization buffer containing 50 mg/ml bovine serum albumin, to give a ,inal concentration of 10 mg/ml BSA. The ba'lls coated with monoclonal antibody were added to the vessel and were incubated four hours on ice 20 with shaking. Labeled antigen was pipetted from the vessel and the balls were rinsed four times with solubilization buffer. The halls were then removed, placed in individual tubes with 100 microliter Laemmli SDS gel sample buffer, and were incubated three minutes in boiling water. The balls were removed and the samples were run on an 25 SDS gel with appropriate standards.
Immunoprecipitation tests on the antibodies indicated that eight of them (2G3, 120H7, 200F9, 204F4, 245E7, 369F10, 7~G6, and 871E3) all bind to high molecular weight mucins (~MW). Two (260F9 and 265B2) bind to the same epitope of a 55 Kd glycoprotein antigen. Two (317G5 and 650E2) bind to a 42 Kd antigen. Two antibodies (451C3 and 454A12) bound to transferrin receptors in the form of a 95 Kd antigen. Neither 451C3 nor 454A12 hlocked binding of transferrin to the receptor. The antigen binding characteristics of the monoclonal antibodles that were tested are summarized in Table 6.
*Trade Mark ; , . . : ~ . ' .. , -~ , . . ' : , ' .. ' ' :,:: . . ' . ~ ' :
~' ~: ' . ' ' ' ' ':

7~7~

ANTIGENS RECOGNIZED BY OVARIA~ MONOCLONAL ANTIBODIES
M M ANTIGEN
___________________________________________________________ _ 2 æ6 75 Kd 3 33F8 66 Kd 44F4 18, 39, 72, 81, 175 Kd (all diffuse bands) 245E7 H~W
11 26~F9 55 Kd 12 266B2 55 Kd 14 317G5 42 Kd 369F10 ~W

17 42lE8 18 451C3 95 Kd (TRANSFERRIN ~ECEPTOU) 19 454~ 2 95 Kd (TRANSFERRIN RECEPTOR) 454C11 200 Kd R 21 650E2 42 Kd 23 ~7lE3 HM~ -Antibody isotype was determ~ned as follows: A grid of 5-mm squares is lightly drawn in pencil on the nitrocellulose sheet and 1-ml droplets of antiisotype sera (Litton Bionetics, Kensington, Maryland, rabbit antisera to mouse ~, ~, a, yl, y2a, y2h, y3, and ~
chains) are applied so that each row of squares receives one spot of each ~heavy and light chain reagent. Ihe sheet is incubated one hour at room temperature in a moist chamber, rinsed quickly in PBS-BSA, containing 1~ ~w/v), and left overnight in PBS~SA at 4C. Strips are cut apart with a scissors and may be stored at 4C in PBS-BSA
containing n.O2% sodium azide. Alternatively, strips may be air-dried and stored desiccated at 4C. A series of small tubes is prepared containing 3 ml hybridoma culture supernatant or supernatant diluted with PBS-BSA, 1:10 dilutions are generally successful; and some supernatants can be diluted as much as 1:200. A nitrocellulose strip is incubated in each tube for one hour at room temperature. Tne .

%~757~

strips are rinsed three times in PBS-BSA and incubated for one hour at room temperature in diluted rabbit anti-mouse-horseradish peroxidase. The strips are rinsed twice in PBS-BSA and twice in Tris buffer. The strips are placed in Tris buffer containing diaminobenzidine and hydrogen peroxide until sufficient color develops on the anti-isotype spots (usually 3-4 minutes). T~e antibody isotypes are indicated in Table 7.

ISOTYPE OF OVARIAN MONOCLONAL ANTIBODIES
MAB ISOTYPE

3 33F~ Gl 12 266B2 Gl ': 14 317G5 G1 16 388D4 Gl 18 451C3 Gl 19 454A12 Gl 22 788G6 Gl , ~ . . .

' ' ' ' :'' ' .'' ':, ~" . -' . . : ~ . .

~37 ~ 7 The antibodies were treated with SPDP as described by ~jorn et al., "Evaluation of ~onoclonal ~ntibodies for the Development of Breast Cancer Immunotoxins," Cancer Res. 45:1214-1221 (19~5) and Carlsson, J. et al., Biochem. J.(1978) 173 723 737 or with iminothiolane (IT) and were conjugated to ricin toxin A chain (RTA) to make the claimed immunotoxins.
SPDP (20 ~M in ethanol) was added in a 20-fold molar excess to antibody and following a 30 min incubation at room temperature, the unreacted SPDP was removed by dialysis against P~S. The extent of deriYatization was determined by measuring the release of pyridine-2-thione at 343 nm after reduction with dithiothreitol (DTT). ~ependiny on the antibody, three to eight lysine amino acid groups (per antihody molecule) were converted to the pyridyl-disulfide derivative.
The SPDP-treated antibodies were conjugated with RT~.
Immediately prior to conjugat;on, the RTA was reduced with 50 mM DTT, then desalted on a column of chromatographic resin containing agarose, dextran and/or acrylamide to remove ~TT from protein. Reduced RTA was added in a three- to five-fold molar excess over pyridyl-disulfide antibody. A typical reaction mixture (1 ml) consisted of 7 ~ antibody and 30 ~m RTA. The reaction was allowed to proceed overnight at 4C. The extent of conjugation of RTA to antibody was determined spectrophotometrically by measuring the release of pyridine-2-thione. On the average, conjugates contained two to three ~TA
molecules per antibody molecule. This was confirmed by nonreducing SDS-PAGE gels (7.5%), which also revealed that the typical conjugate preparatlon contained 10% 30~ free antibody.
The cnnjugate mixture was chromatographed on a HPLC size exclusion column to separate conjugates from residual unreacted RT~.
The column was equilibrated in 0.1 sodium sulfate/0.02 M sodiuln phosphte pH 6.8. Conjugate mixture (~.7 ml) was injected, then chromatographed at a flow rate of 1 ml/min (room temperature).
Fractions of 0.5 ml were collected and the peak conjugate fractions were pooled and filter sterilized prior to cytotoxicity testing.

. .
~ . . - ~ .
.
., , .... .. . -, , :
~-,, ., . . ~. :. . , ;,- . ,~ . :, . :
.

'.7 ` ` ,i ~L~ ~ 7 ( -25-Approximately 30 mg/ml antibody in O.I0 M Na phosphatet 0.001 M Na EDTA, pH ~.0 (hereafter referred to as P~DTA buffer) is reacted with 1 mM 5,5'-dithiobis-( 2 -nitrobenzoic acid) (DT~B) at room temperature for about 15 min and then ch111ed to 0C ln an ice bath.
Enough IT is added to this solution to give an average of 2.5 IT
molecules/antibody molecule, and the resulting solution is dialysed at 0-5C against three 100-fold excess volumes of P-EDTA buffer.
RTA, normally stored in P~DTA containing 1 mM DTT, is ultrafiltered to a concentration between I0 and 15 mg/ml and dialyzed at 0-5C against three 100-fold excess volumes of P~DTA. Enough RTA
is added to the derivatized antibody to give an average of 1.0-1.2 Free thiols on RTA per blocked thiol on derivatized antibody. Tnis mixture is incubated at room temperature for 2 hrs.
The coupling reaction mixture is applied to a column of a chromatographic resin based on a blue dye (trysacryl blue) covalently coupled to a solid support, which mixture is then eluted with P~UTA
at room temperature. T~e column is scaled to contain approximatel~Y 2 ml of bed volume per mg of starting antibody, After an initial peak~
of unconjugated antibody has been eluted from the column, the elutant is switched to P-EDTA containing 1 ~ ~aCl. Immunoconjugate and '"` unreacted RTA are eluted in this buffer as a very s~arp peak, which is pooled and dialyzed at 0-5C against one 10-fold excess volume of 0.15 M Na phosphate, pH 7.1 (hereafter referred to as pj buffer). The dialyzed protein is applied to a column of a size-exclusion gel at 0-5C and eluted with buffer at a flow rate of 6 cm/hr. The column is scaled to contain at least 25 ml of bed volume/ml of applied protein. Immunoconjugate is eluted as a single peak, slightly after the excluded volume, baseline-resolved from following peaks of dimerized and monomeric RTA. The pooled immunoconjugate peak is ultrafiltered at 35 psi to a final concentration of 5.0 mg/ml and filter~sterilized. -~
The invention wiil be better understood in light of the following examples which are intended by the inventor to be merely exemplary and non-limiting.

.
., . ~ ., , . :
,, ~ ',.
- . ' :~. . .
,' , . .. . . ..

~7~

Example I
Female athymic nude mice (Nu/Nu~ strain ~alh/C), weighing between 16 and 22 grams were used. NIH:OVCAR-3 asc~tes cells were ohtained from carrier micè. The cells were washed twice in phosphate buffered saline (PBS) and resuspended in P~S at approximately 1 volume of cells to 2 volumes of PBS. Cell count was determined by counting in a haemocytometer. Cell viability was determined by trypan hluP dye exclusion. Each animal was injected intraperitoneally with 5x107 viable cells on day zero. Animals were injected with the immunotoxins on days 4, 7 and 10. The immunotoxin was usually admistered in 0.1ml `~ P8S. Con~rol animals were injected with 0.1ml PBS on the same schedule. Five animals were used for each dose of each immunotoxin tested and for the controls. Animals were observed daily.
Effectiveness was determined by an increase in survival time relative to controls in each experiment or by less abdominal swelling due to retardation of the increase in tumor burden in treated animals as compared to controls having the same survival time.
The results are reported in Table 8. In Tahle 8, and the~
following tables, "Swelling Index" is defined as follows: 0 = no abdominal distension; 1 = barely visible abdominal distension; 2 =
moderate abdominal distension; and 3 = severe abdominal distension.

. , .
: - ' ~: ;,, . '' - . . .. , : , . .
. . . . .: .
: ~ '' "' ,. ' ., .

7~8 ( Experiment A
Test ~aterial Dose #Surviving Swelling Index ~ean Life Span ________ _ __ _ __~ _ ___ _ _ __ ____ __ _ _ _ __ _ _ __ _ 317G5-1T-RTA 50ug 0 - 49-~ +/- 10 100ug 2 3 60.2 ~/- 5.2 260F9-IT-~TA 50ug 0 - 26 +/- 1.4 100ug 0 - 24.6 ~/- 3.3 113F1~IT-RTA 25ug 0 3 32.2 ~/-13.9 50ug 0 - 29 ~/-3.0 P~S Controls 0.1ml 0 - 29 Experiment B
Material Dose #Surviving Swelling Index ~ean Survival (day 85) _____ _ ________________ _____ ________ ____ _ ____ _.. _____ _ 454A12-IT-rRTA 25ug 1 2 >74 ` 280D11-IT-RTA 50ug 1 2 >66 100ug 2 2 >71 i 2G3-IT-RTA 50ug 0 3 30 oou9 n 3 35 Example II
In the following example the experiment was run essentially as described in the previous example except that the animals were injected on days 4, 6 and 8. This example shows that the anti-tumor effect of immunotoxin 454 ~ 2-IT-rRTA is blocked when tumor bearing animals are treated with an excess of the monoclonal antibody 454A12 from which the immunotoxin is derived. ~OPC21, an antibody which is not human ovarian tumor specific, when administered at excess with 454 ~ 2-lT-rRTA has no corresponding blocking effect.

- : ~ . . .: .
- . . .
- .
. . . . .
.

12875*~
! -28-Material Dose(ug)~ # Surviving Swelling Survival (day 69) Index (mean days) _________ __, _ _ __ _ _ _ .. __ _____ __ _ ___ ___ _ _ 454A12-IT-RTA 25 4 0 >69 454~2-ITRTA 25 0 3 26 +454A12(500ug) 454A12-IT-RTA 25 3 1 >65 ~OPC21 (500ug) 317G5-IT-RTA 50 2 0 >60 100 4 0 >65 Exa~ple III
The procedure used in this experiment is essentially the same as Exampl e I . This experiment shows that immunotoxins co~prised of the Fab'2 fragment of 454A12 conjugated to RTA has antitumor activity comparable to 454A12-ITRTA.

Material Dose(ug)# # Surviving Swelling Survival (day 34) (mean days) .
__ _______ __ _ _ __ ______ _ _ ___ _____ __ PBS
454A12-IT-RTA 10 2 0 >34 454A12-IT-RTA 25 3 0 >39 454A12-IT~TA 50 4 0 ~39 454A12-RTA 10 3 0 >39 (Fab'2) 25 4 0 >39 . 50 3 0 >34 : .:

- : . . . ... . ..

: .; .

Example IV
The followiny Example shows the in vitro cytotoxicity of the ~mmunoconjugates against several ovarian cancer cell lines.
~ IH:OVCAR-2, -3, -~, an~ -5 are isolates f rom the malignant ascites of patients with ovarian carcinoma. ~ ese cell lines have been previously described in the following references which are herein incorporated by reference. Hamilton et al., "Cnaracterization of Human Ovarian Carcinoma Cell Lines (NIH:OVCAR-3) with Androgen and Estrogen Receptors" Cancer Res 43:5379-5389 (1983). Hamilton et al,, "Experimental ~odel Systems of Ovarian Cancer: Aplications to the Desjgn and Evaluation of New Treatment ~pproaches" Seminars in Oncolo~y 11:285-298 (1985). The ovarian cancer cell line A1~47 was obtained from S. ~ronson (National Cancer Institute, Bethesda, ~aryland). The ovarian cells were grown in RPMI medium 1640, 10%
fetal bovine serum, 10 ;g/ml insulin and penicillin-streptomycin. KB
cells were grown in Dulbecco's Modified Eagle Medium (DME~), 10% calf serum, glutamine and penicillin-streptomycin. Tissue culture media, sera, glutamine and antibiotics were purchased from Grand Island Bio10gical Col, Grand Island, NY, and insulin was obtained from Elanco 20 Products Company~ Indianapo1is, IN. For protein synthesis inhibition ~`~` assays, cells were plated at 2 x 105 cells/35-mm dish one day prior to use. ~efore adding imnlnotoxins, cells were washe~ twice with n~E~
containing bovine serum albumin (2 mg/ml) (o~EM~SA). The listed imm~notoxins were made by iminothiolane derivitization and conjugation 25 to RTA as described hereinabove.
~ Inhibition of protein synthesis was used to measure the activity of the immunotoxins~ Cells were incu~ated with DMEM ~SA
containing various concentations of immunotoxins at 37C for 24 h and then assayed for incorporation of [3H]leucine (New England Nuclear, Boston, MA; specific activity 140.8 Ci/mmol) into TCA-insoluhle material as described in Pirker et al. "~nti-Transferrin Receptor Antibody Linked to Pseudomonas Exotoxin: A Model Immunotoxin in Human Ovarian Carcinoma Cell Lines" Cancer 45:751-757 (1985). ~ean values of duplicates were expressed as a percentage of controls of the same . ~- . . ~ .
, . .
- . .
, .-, .
.. - ~ ~ . . . . . ..
, . . . ... .
. . - . , `~ ~ 2 ~7 ~8 cell line which did not receive immunotoxins. Immunoconjugates that gave 50% inhibition of protein synthesis as conpared to untreate-l controls (ID50) of 10nM or less were considered to be effective. I~50 of the immunoconjugates tested are listed below in Table 11.

IN VITRO CYTOTOXICITY
I~50(n~
RTA CONJUGATF OV-2 OV-3 OV-4 OV-5 ~847 KB
__ 4S4A12 0.04 0.05 0.05 0.03 --- 0.01 317G5 0.1 0.2 0.1 0.3 --- 0.1-2 260F9 0.2 0.5 0.2 0.2 >5 140 113Fl -^- 2 --- --- --- ---280Dll >30 4 13 >20 >30 120 454C11 >s >5 >5 >5 ~5 >5 520C9 >5 ___ ___ ___ __ _ 245E7 >30 >30 >30 30 >30 ~30 .
Examole V
The immunoconjugates described in the immediately preceeding example were tested against NIH:OVCAR-3 cells. Cells were maintained in RPM1 medium 1640, 10% fetal bovine serum, 10 lg/ml insulin and penicillin-streptomycin. Cells were removed from the culture flasks by mild trypsin digestion or versene addition. T~e cell concentration was adjusted. 4 x 105 NI~:OVCAR-3 cells were sl~spended in 1 ml of medium and were added to 8-ml glass vials (ICN), followed by the addition of conjugate dilutions (in PBS containing bovine serum albumin, 100 !Ig/ml), After incubation at 37 for 22 ~rs., the medium was aspirated, the monolayers were washed with PBS, and 0.5 ml methionine-free medium supplemented with 0.5 ~Ci L-[35S]methionine (Amersham; 1400 Ci/mmol) was added. After a 2-hr incubation at 37, the~ medium was aspirated, and the cell monolayers were washed twice ,,. , " ,. ~ . ... .
- . . ~ . . , ' -: : . . .

.' . ~ ' ',:
.
.

~287S7~

( -31-with 10% trichloroacetic acid containing methionine (1 mg/ml). The cells were dried, scintillation fllJid was added, and the radioactivity was measured in a Packard CL/D liquid scintillation counter.
Inhibition of protein synkhesis was calculated as the incorporation of TCA precipitable 35S counts for each vial. ~ean values were expressed as a percentage of controls of the same cell line that did not receive immunotoxins, ID50's were determined as in the immediately preceeding example. The results are reported in the following Table 12.

.

:

~: :

:

:

:
.

,, ~ . .. . . .

, ~ ,, , , , : .,:
: . ~ . ,.. . :, . : ~ , -, . ,. ., ~ - .

lZ87~;i7B

In Vitro Cytotoxicity vs O~CAR-3 -CONJ~GATE ID5n(n~) 454A12 RTA 0.05 454A12-RTA 0.2 454A12-(Fab')2 RTA 0.4 317G5-RTA 0.2 2G3~ TA 3 260F9~TA 4 369F10-RTA >56 245E7 ~TA >56 520C9-RTA >112 ~OPC21RTA >112 MOPC21-RTA >80 Examp~ VI
This Example shows the cytotoxicity of immunoconjugates comprising the monoclonal antibodies described above and Pseudomonas exotoxin.
Pseudomonas exotoxin (PE) was a gift of Dr. S. Leppla (Ft.
Detrick, Frederick, ~D). PE may also be ohtained commercially fro~
Swiss Serum and Vaccine Institute, Berne, Switzerland. PE conjugates were constructed and purified by a modification of a method previously disclosed. Pi rker et al. (1985). PE (30 nmol) was reacted with 5000 nmol 2-iminothiolane-HC1 (Pierce Chemical Co., Rockford, IL) and ~500 nmol NAD+ in 1 ml ~.1 M phosphate buffer (pH
8.0)~ containing 1 ~ EGTA at 37C for 1 h. The derivatized PE was then~separated fram the reactants using HPLC and activated by the addition~of~5~,5~'-dithio-bis(2-nitrobenzoic acid) (~TNB) to a final 15~concent~ration of 1 ~M. Antibodies (40-50 nmnl) were incubated with 100-200 nmol~ 2-iminothiolane-HCl in 0.75 ml 0.1 ~ phosphate buffer (pH
8.0) containing ~1 ~M EGTA at 37C for 1 h. The antibodies were react~ed wlth the activated PE and the conjugates were purified using -:

. . .

- ,,~ . :, ~' ' ' ' .' ., ' .. . .. . ,~ .
~, : ~ ' -' , ' ' ' . : :
. .

3L2~7S7~
I'' HPLC as described. Pirker et al. (19~5). A peak containing a on~to-one conjugate of PE with the antihody was recovered and used for all studies described below.
Inhibition of protein synthesis and ID50ls were determined as descrihed above in Example IV except that the cells were incubated with immunotoxin for 12 ~rs. Results from representative protein inhibition assays are shown and the average ID50 values of all experiments are provided in Table 13. ID50.s are shown as ng/ml and (n~) in the table.
`~; `

, .

:~ , ::

.

., - .

. ~ ~ , . , , :

2~ S 7 ID50-Values in rg/ml (nM) for Protein Synthesis Inhibitiona Cells 454C11-PE 260F9-PE 280D11-PE
_______ _______ ____ ___ ___ _________________ ____ __ _ __ ____ __ OVCAR-2 1.6(0~01) 3~4(0.02) 835(4) OVCAR-3 3.6(0.02) 41.5(0.2) 805(4) OVCAR-4 0.7(0.005) 4.7(0.02) 54(~.3) OVCAR-5 10(0.05) 23(0.1) 3450(>15) A1847 2.5(0.015) 385C(2) 2200~10) KB lSb(0.08) >600(>3) >250(>1) a If not otherwise mentioned, these values are mean values of at least two experiments.
b Results from one experiment.
c Non-specific toxicity.

Samples of the hybridomas that produce the monoclonal antibodies from which the immunotoxins according to the invention are derived have been deposited in the American Type Culture Collection or j.~
the Collections of In Vitro International under the following accession numbers:
ATCC
HybridomaAccession No.

280~ 8487 266B2H~ 8486 245E7HB 84~9 .

, . ~ . ,', ,...... . .: . .
:

~ lZ~57~

In Vitro International Collection Hybridoma ~cession No.

87 lE3 ~` ` 650E2 IVI 10083 These deposits were made under the Budapest Treaty and will be maintained and made accessible in accordance with the provisions thereof.

:

~: :

, -:~ :

: ~' :::

.. . .: , - . . .

Claims (11)

1. Immunotoxin comprising a cytotoxic moiety and an antigen binding portion selected from the group consisting of the Fab, Fab' and F(ab')2 region of a monoclonal antibody wherein said monoclonal antibody (i) binds human ovarian cancer tissue;
(ii) has a selectivity of about 0.11 or less;
(iii) is an IgG or IgM
said immunotoxin having at least one capability selected from the group consisting of:
a cytotoxicity ID50 of about 10 nM or less against human ovarian cancer cells; retarding the rate of growth of tumors comprised of human ovarian cells carried by a mammal when said mammal is treated with said immunotoxin; or extending the survival time of a mammal bearing a tumor comprised of human ovarian cancer cells when said mammal is treated with said immunotoxin.

2. The immunotoxin of claim 1 wherein the human ovarian cancer cells are at least one selected from the groups consisting of OVCAR-2, OVCAR-3, OVCAR-4, OVCAR-5 and A1847.

3. The immunotoxin of claim 1 wherein said monoclonal is selected from the groups consisting of 2G3, 9C6, 33F8, 44B2, 44F4, 120H7, 200F9, 204F4, 219F3, 245E7, 260F9, 266B2, 280D11, 317G5, 369F10, 388D4, 421E8, 454C11 454A12, 451G3, 650E2, 788G6, 871E3 and monoclonal antibodies that are the functionally equivalent to a number of said groups.

4. The immunotoxin of claim 1 wherein said monoclonal antibody binds an antigen selected from the group consisting of a high molocular weight mucin, one epitope of a 55 Kd antigen which can be bound by 260F9 or 266B2, a 200 Rd antigen, and a 42 Kd proteinaceous antigen.

5. The immunotoxin of claim 1 wherein the toxic moiety is an enzymatically active toxin of bacterial, plant or fungal origin, selected from the group consisting of ricin toxin A chain, Phytolacca americana proteins, diphtheria toxin A fragment, non-binding active fragments of diphtheria toxin A fragment and Pseudomonas aeruginosa exotoxin A.

6. The immunotoxin of claim 1 wherein the ricin toxin A chain is recombinant ricin toxin A chain.

7. Immonotoxin of claims 1 comprising at least an antigen binding portion of a monoclonal antibody wherein said monoclonal antibody binds to human transferrin receptor but does not block the binding of transferrin thereto.

8. The immunotoxin of claim 7 wherein said antigen binding portion of a monoclonal antibody comprises the F(ab')2 portion thereof.

9. For use in extending the survival time of a mammal bearing tumors comprised of human ovarian tumor cells, an amount of an immunotoxin of claims 1, 2 or 7 effective to extend the survival time of said mammal.

10. For use in retarding the rate of growth of tumors comprised of human ovarian cancer cells carried by a mammal, an amount of an immunotoxin of claims 1, 2, or 7 effective to retard the rate of growth of human ovarian tumors carried by said mammal.

11. For use in the killing of human ovarian cancer cells, a cytotoxically effective amount of an immunotoxin of claims 1, 2, or 7.